Hipster Wars: Discovering Elements of Fashion Styles

Abstract

The clothing we wear and our identities are closely tied, revealing to the world clues about our wealth, occupation, and socioidentity. In this paper we examine questions related to what our clothing reveals about our personal style. We first design an online competitive Style Rating Game called Hipster Wars to crowd source reliable human judgments of style. We use this game to collect a new dataset of clothing outfits with associated style ratings for 5 style categories: hipster, bohemian, pinup, preppy, and goth. Next, we train models for betweenclass and within-class classification of styles. Finally, we explore methods to identify clothing elements that are generally discriminative for a style, and methods for identifying items in a particular outfit that may indicate a style.

Interleaved text/image Deep Mining on a large-scale radiology database

“…Low-level topic images are visually more coherent (i.e., may be easier to learn). High-level topics may be analogous to [21,34]. About half of the key images are associated with topic #77, implying that the clinicians' image referencing behavior patterns heavily focuses on metastatic patients.…”

Multi-modal Embedding for Main Product Detection in Fashion

“…Common examples are clothing classification or retrieval [3,2,23], clothing parsing (i.e., semantically label each image pixel) [29,21,4,15] or higher level tasks such as evaluating style or deducting people's occupation or social tribe [24,11,28,22,17].…”

Discriminative learning of apparel features

“…Due to the importance of the market for online shopping of dressed clothing, various methods have been recently proposed to tackle the problem of classification of apparel images [11,12,13,14,15]. Most of them work with images taken in unconstrained environments and try to classify the apparel [11,13], retrieve related images in the database [12] or infer how much a person displays a certain style [14].…”

“…Most of them work with images taken in unconstrained environments and try to classify the apparel [11,13], retrieve related images in the database [12] or infer how much a person displays a certain style [14].…”